Hardware Configs¶

Which setup are you?¶

Pick the row that matches your sensors. Everything else follows from there.

My sensors	Use case	Start here
IMU + wheel odometry	Indoor, no GPS: warehouse, home robot, lab	`wheels_indoor.yaml`
IMU + LiDAR ICP	Indoor, no wheel odom: using KISS-ICP or rtabmap	`icp_indoor.yaml`
IMU + wheel odom + LiDAR ICP	Indoor, both velocity sources for redundancy	`wheels_indoor.yaml` + uncomment `encoder2.topic`
IMU + wheel odom + GPS	Outdoor, field robots, GPS navigation	`bno085_custom.yaml` or `clearpath_husky.yaml`
IMU + VESC encoder + RealSense D435i, Ackermann	F1/10 scale car, indoor	`f1tenth_indoor.yaml`
IMU + ZED 2i visual odom + GPS, Ackermann	Van or large vehicle, outdoor	`van_outdoor_gps.yaml`
IMU + VSLAM (ORB-SLAM3, RTAB-Map, Kimera)	Indoor, GPS-denied, no wheel odom	`vslam_imu.yaml`

Variations within each setup: one-line changes, not separate configs:

Variation	Parameter to change
Standard GPS → RTK float	`gnss.base_noise_xy: 0.5`, `gnss.min_fix_type: 2`
Standard GPS → RTK fixed	`gnss.base_noise_xy: 0.015`, `gnss.min_fix_type: 4`
NavSatFix → GPSFix (RTK_FLOAT status, native HDOP/VDOP)	`gnss.use_gps_fix: true` (requires driver that publishes `gps_msgs/GPSFix`)
6-axis IMU → 9-axis (BNO085, VectorNav)	`imu.has_magnetometer: true`
Add dual GPS antenna heading	`gnss.heading_topic: "/gnss/heading"`
Add a second IMU (dual IMU redundancy)	`imu2.topic: "/your/second/imu"`
Add Nav2	Use `fusioncore_nav2.launch.py` instead of `fusioncore.launch.py`
Urban vs open sky GPS noise	Layer an environment preset on top

All config files¶

Config	Setup	Platform	IMU	GPS
`wheels_indoor.yaml`	IMU + wheels	Any differential drive (generic)	Any 6-axis MEMS	none
`icp_indoor.yaml`	IMU + LiDAR ICP	Any indoor robot with LiDAR	Any 6-axis MEMS	none
`bno085_custom.yaml`	IMU + wheels + GPS	Any differential drive (DIY)	Bosch BNO085	u-blox M8N class
`clearpath_husky.yaml`	IMU + wheels + GPS	Clearpath Husky A200	Microstrain 3DM-GX5-25	u-blox F9P
`duatic_mecanum.yaml`	IMU + wheels	Duatic mecanum platform	BNO085	none
`f1tenth_indoor.yaml`	IMU + VESC encoder, Ackermann	F1/10 scale car (indoor)	RealSense D435i	none
`van_outdoor_gps.yaml`	IMU + visual odom + GPS, Ackermann	Full-size van / large vehicle	ZED 2i	any NavSatFix
`vslam_imu.yaml`	IMU + VSLAM pose	Any robot with visual SLAM, indoor GPS-denied	VectorNav / 9-axis	none

Using a config¶

ros2 launch fusioncore_ros fusioncore.launch.py \
  fusioncore_config:=$(ros2 pkg prefix fusioncore_ros)/share/fusioncore_ros/config/wheels_indoor.yaml

Add remaps if your topics differ from the defaults:

ros2 launch fusioncore_ros fusioncore.launch.py \
  fusioncore_config:=$(ros2 pkg prefix fusioncore_ros)/share/fusioncore_ros/config/wheels_indoor.yaml \
  --ros-args \
  -r /odom/wheels:=/diff_controller/odom \
  -r /imu/data:=/imu

Indoor / GPS-denied robots¶

See Indoor / LiDAR ICP for detailed setup instructions: KISS-ICP, rtabmap, OAK-D IMU remaps, SLAM integration.

See VSLAM + IMU for ORB-SLAM3, RTAB-Map, Kimera, and other visual SLAM systems: covariance requirements, VectorNav setup, ORB-SLAM3 reinitialization handling.

Ackermann / car-like robots¶

See Ackermann vehicles for F1/10 racers, vans, and anything with front-wheel steering: wheelbase measurement, VESC covariance handling, ZED 2i gravity removal, multiple IMU guidance, and publish.tf.

Layering environment presets¶

See Environment Presets to tune GPS noise for your operating conditions without editing the hardware config.

Don't see your hardware?¶

Open a Hardware Config Request with your sensor datasheets and the community will help put one together.